Aim. Bone and joint infections (BJI) need frequently prolonged antibiotic treatment at high dosage for a total of 6 or 12 weeks depending the type of infection. Impact of such prolonged antibiotic exposure on the gut microbiota has never been assessed. Method. We performed a national multicentric prospective study of patients with BJI to monitor the gut microbiota dynamic all along antimicrobial treatment. Clinical data and stool collection were performed at the baseline visit (B) within 24h before starting antibiotics, at the end of the treatment (EOT) and 2 weeks after antibiotic withdrawal during a follow-up visit (FU). Microbiota composition was determined by shotgun metagenomic sequencing. Biological markers of gut permeability and inflammation were monitored at each time point. Results. Sixty-two patients were enrolled: 27 native BJI, 14 osteosynthesis-related BJI and 21 prosthetic joint infections (PJI). At EOT there was a significant loss of alpha-diversity that recovered at FU in patients with native BJI and PJI but not in patients with osteosynthesis-related BJI (p<0.05, Wilcoxon test). At EOT, we observed an increase of Proteobacteria and Bacteroidetes that partially recovered at FU. Principal Component Analysis (PCoA) of the Bray Curtis distance, showed a significant change of the gut microbiota at the end of treatment compared to baseline (p<0.01, PERMANOVA) that only partially recover at FU. The taxonomic analysis showed that microbiota composition at FU does not differ significantly at the genus level when comparing patients treated for 6 weeks to patients treated for 12 weeks. No particular antibiotic (especially fluoroquinolones) was associated with a lower Shannon index or distinct dynamic of recovery at the end of treatment. PCoA analysis of the Bray Curtis distance shows that patients with elevated plasma level of CRP (≥5mg/L) at EOT had a distinct gut microbial composition compared to others. Conclusions. In patients with BJI, antibiotics altered the gut microbiota diversity and composition with only partial recovery 2 weeks after antibiotic withdrawal, independently on the duration of the therapy and on the type of the antibiotic used. Elevated CRP at EOT might reflect persistent alteration of the gut microbiota. Assessment of long-term impact after the end of treatment is on-going